CN113328881A - Topology sensing method, device and system for non-cooperative wireless network - Google Patents

Topology sensing method, device and system for non-cooperative wireless network Download PDF

Info

Publication number
CN113328881A
CN113328881A CN202110576245.9A CN202110576245A CN113328881A CN 113328881 A CN113328881 A CN 113328881A CN 202110576245 A CN202110576245 A CN 202110576245A CN 113328881 A CN113328881 A CN 113328881A
Authority
CN
China
Prior art keywords
data
topology
node
hypothesis
test
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110576245.9A
Other languages
Chinese (zh)
Other versions
CN113328881B (en
Inventor
刘子彤
王威
吴启晖
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing University of Aeronautics and Astronautics
Original Assignee
Nanjing University of Aeronautics and Astronautics
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nanjing University of Aeronautics and Astronautics filed Critical Nanjing University of Aeronautics and Astronautics
Priority to CN202110576245.9A priority Critical patent/CN113328881B/en
Publication of CN113328881A publication Critical patent/CN113328881A/en
Application granted granted Critical
Publication of CN113328881B publication Critical patent/CN113328881B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/12—Discovery or management of network topologies
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/142—Network analysis or design using statistical or mathematical methods
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • Algebra (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Physics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

The invention discloses a topology sensing method, a device and a system facing a non-cooperative wireless network, wherein the method comprises the following steps: acquiring a node serial number and a sending moment of data sent by a target network within a period of time through a signal detection mechanism to form a data matrix; performing Glan's cause and effect hypothesis test on the data matrixes pairwise, solving the mean value of Glan's cause and effect zero distribution by using a time window method, and screening to obtain a potential neighbor set of the nodes as a threshold value; and carrying out conditional granger causal hypothesis test according to the data matrix and the potential neighbor set group, solving the mean value of conditional granger causal zero distribution by using a time window method, and screening as a threshold to obtain a final neighbor set of the node to realize topology perception. By the topology sensing algorithm based on the conditional granger cause and effect, the communication relation can be inferred according to the signal sending data matrix under the condition of not decoding the data packet, the network topology information can be accurately inferred, and the topology inference can be carried out by utilizing the perception information.

Description

Topology sensing method, device and system for non-cooperative wireless network
Technical Field
The invention belongs to the technical field of communication, and particularly relates to a topology sensing method, a device and a system for a non-cooperative wireless network.
Background
The topology perception means that the topological structure of the network is inferred through perception of information such as frequency spectrum data and communication time sequence of a target network, so that subsequent battlefield decision is guided. The network topology information is an extremely important network high-level semantic reasoning, the network topology of an enemy is accurately identified on a battlefield, the overall situation of the enemy network can be known, and strategic decision implementation is facilitated; meanwhile, key nodes and key links of the network can be further identified on the basis of the network topology, so that accurate interference can be carried out on the network, or battlefield resource management decision can be assisted.
However, non-partner-oriented wireless network topology awareness presents significant challenges. Reasoning about the network topology of the non-cooperator and reasoning about the network topology of the own partner are different and more difficult. This is because the non-cooperative parties result in limited and unreliable information available to the perceiver, and the inability to control the target network data flow also means that the related methods based on network protocols do not work. Therefore, it is urgent to find an effective topology sensing method oriented to a non-cooperative wireless network.
Disclosure of Invention
The technical problem to be solved by the present invention is to provide a topology sensing method, device and system for a non-cooperative wireless network, aiming at the deficiencies of the prior art.
In order to achieve the technical purpose, the technical scheme adopted by the invention is as follows:
a topology sensing method facing a non-cooperative wireless network detects the relevance of nodes according to the sending time of signals and the serial number of the sending node, analyzes the communication relation of a target network, and accordingly obtains a topology structure, and specifically comprises the following steps:
step 1: acquiring a node serial number and a sending moment of data sent by a target network within a period of time through a signal detection mechanism to form a data matrix P;
step 2: performing Glan's cause and effect hypothesis test on the data matrixes pairwise, solving the mean value of Glan's cause and effect zero distribution by using a time window method, and screening to obtain a potential neighbor set of the nodes as a threshold value;
and step 3: and carrying out conditional granger causal hypothesis test according to the data matrix and the potential neighbor set group, solving the mean value of conditional granger causal zero distribution by using a time window method, and screening as a threshold to obtain a final neighbor set of the node to realize topology perception.
In order to optimize the technical scheme, the specific measures adopted further comprise:
the step 1 is specifically as follows:
the sensor compares the received signal energy value with a signal detection threshold by sampling to determine if a signal is present:
if the signal exists, discretizing the signal according to the time slot by taking the falling edge of the signal as a standard, as shown in the following formula,
Figure BDA0003084445310000021
wherein xi,lData value, T, representing the ith time slot of the ith nodesIndicating the time length of the time slot, tiRepresenting the signal arrival time of the ith node;
thereby obtaining a preprocessed data matrix P
Figure BDA0003084445310000022
The step 2 mentioned above performs glange causal hypothesis test on the data in the data matrix two by two rows, including setting zero hypothesis and alternative hypothesis, and calculating statistical test quantity, and the specific steps are as follows:
step 2.1.1: data was modeled as the grand causal hypothesis:
for variables X and Y, zero hypothesis H0：
Assuming that without prior information for the variable Y, the information for X can be well predicted, i.e. Y cannot cause X by glange;
alternative hypothesis H1：
Assuming that the a priori information of Y helps the prediction of the information of X, i.e. Y granger causes X, it is expressed by the following equation:
Figure BDA0003084445310000023
Figure BDA0003084445310000024
wherein X [ n ]]Is the value of X in the nth slot, aiAnd biIs a parameter of linear regression, quantifies the degree of influence of past values on the current value, k is the order of the model, ε [ n ]]And eta n]Are each H0And H1An error of (2);
step 2.1.2: calculating a statistical test quantity G:
Figure BDA0003084445310000031
wherein
Figure BDA0003084445310000032
Denotes a null hypothesis H0The sum of the squares of the residuals of (c),
Figure BDA0003084445310000033
representing alternative hypothesis H1The sum of the squared residuals of (c). k denotes the order of the model, M denotes the number of observation total slots, F (d)1,d2) Denotes F distribution, d1＝k，d2＝M-2k-1。
The method of using the time window in the step 2 obtains the average value of the granger causal zero distribution, and obtains the potential neighbor set of the node as the threshold value screening, specifically as follows:
step 2.2.1: selecting a proper time window number nwDividing the data into n with a time windowwThen calculating the granger causal test statistic G;
step 2.2.2: calculating the test statistic of all nodes, then calculating the mean value, screening the test statistic, considering that the two variables are related when the mean value is larger than the threshold value, and considering that the two variables are unrelated when the mean value is not larger than the threshold value; the relevant variables are then put together as a set of potential neighbors for the node.
The conditional grand cause and effect hypothesis test is performed according to the data matrix and the potential neighbor set grouping in the step 3, and the specific steps are as follows:
step 3.1.1: in a neighbor set K of a node X, respectively checking whether each node has a special effect on the information prediction of the node X;
null hypothesis S0: all node information except Y in the known K set plays a role in predicting node X information, namely Y plays no special role in X;
alternative hypothesis S1: compared with the null hypothesis, the prior information of Y has a role in predicting X information, namely Y has a special role in X and is expressed by the following formula:
Figure BDA0003084445310000034
Figure BDA0003084445310000035
wherein X [ n ]]Is the value of X in the nth slot, ai,biAnd eiIs a parameter of linear regression, so as to quantify the degree of influence of the past value on the present value, k is the order of the model, ε [ n ]]And eta n]Are each H0And H1An error of (2);
step 3.1.2: calculating the statistical test quantity Gc：
Figure BDA0003084445310000041
Wherein
Figure BDA0003084445310000042
Denotes a null hypothesis H0The sum of the squares of the residuals of (c),
Figure BDA0003084445310000043
representing alternative hypothesis H1The sum of the squared residuals of (c). k denotes the order of the model, M denotes the number of observation total slots, F (d)1,d2) Denotes F distribution, d1＝k，d2＝M-2k-1。
The method of using the time window in the step 3 obtains the average value of conditional grand causal zero distribution, and obtains the final neighbor set of the node as the threshold value screening, and the specific steps are as follows:
step 3.2.1: selecting a suitable number of time windows, e.g. nwData were divided into two parts with a time window and the conditional granger causal test statistic G was calculatedc. For example, for variables X and Y, they are classified as X1，X2，Y1，Y2Then separately calculate X2For Y1、Y2To X1Test statistic Gc；
Step 3.2.2: calculating the test statistic of all nodes, then calculating the mean value, screening the test statistic, considering that the two variables are related when the mean value is larger than the threshold value, and considering that the two variables are unrelated when the mean value is not larger than the threshold value; the relevant variables are then put together as a set of potential neighbors for the node.
A topology aware apparatus for non-cooperative wireless networks, comprising:
a data preprocessing module: the data matrix P is used for preprocessing the sensed data so as to obtain a data matrix P as initial data of topology inference;
a potential neighbor identification module: performing the granger causal hypothesis test by using the data provided by the data preprocessing module, calculating test statistic, and performing potential neighbor screening by using the mean value of zero distribution as a threshold value;
and a final neighbor identification module: and carrying out conditional granger causal hypothesis test by using the processing result of the potential neighbor identification module, calculating statistic, and carrying out final neighbor screening by using the mean value of zero distribution as a threshold value.
The data preprocessing module comprises:
an initialization unit: initializing the device and emptying the data in the previous time period;
a sensing unit: the method comprises the steps of sensing signals, and obtaining signal sending time and node serial numbers;
a processing unit: the signal discretization method comprises the steps of discretizing a signal by taking a signal falling edge as a standard;
a storage unit: the data processing device is used for storing the processed data;
an output unit: for outputting the processed data to the potential neighbor identification module and the final neighbor identification module.
The potential neighbor identification module and the final neighbor identification module both include:
statistic calculation unit: test statistics for calculating a granger causal hypothesis test;
a zero distribution calculation unit: for calculating the mean of the granger causal zero distribution;
the potential neighbor identification module further comprises: potential neighbor screening unit: for screening a set of potential neighbors;
the final neighbor identification module further comprises: and a final neighbor screening unit: for screening the final set of neighbors.
A topology aware system for non-cooperative wireless networks, comprising:
a memory for storing instructions and data;
a processor coupled to the memory, the processor configured to invoke and execute instructions and data stored in the memory, specifically:
the processor is used for preprocessing the sensed data to obtain a data matrix P which is used as initial data of topology inference; the method is used for carrying out the Glange causal hypothesis test, calculating test statistics, and carrying out potential neighbor screening by taking the mean value of zero distribution as a threshold value; the method is used for carrying out conditional grand causal hypothesis test, calculating statistic, carrying out final neighbor screening by taking the mean value of zero distribution as a threshold value, and storing the threshold value in a memory.
The invention has the following beneficial effects:
by the topology sensing algorithm based on the conditional granger cause and effect, the communication relation can be deduced according to the signal sending data matrix under the condition of not decoding the data packet, and the network topology information can be accurately deduced; the method can utilize the perception information to carry out topology reasoning in a plurality of tasks such as network countermeasure, military investigation, network management and the like, thereby providing a basis for subsequent resource allocation and decision making.
Drawings
FIG. 1 is a schematic diagram of an implementation scenario of the present invention;
FIG. 2 is a schematic flow chart of a topology sensing method for a non-cooperative wireless network according to the present invention;
FIG. 3 is a flow chart of the conditional Graandy causal topology awareness algorithm of the present invention;
FIG. 4 is a schematic structural diagram of a topology sensing device facing a non-cooperative wireless network according to the present invention;
fig. 5 is a schematic structural diagram of a topology awareness system oriented to a non-cooperative wireless network according to the present invention.
Detailed Description
Embodiments of the present invention are described in further detail below with reference to the accompanying drawings.
Referring to fig. 1, an implementation scenario of the present invention relates to a target network and a sensing network, wherein the target network is composed of a plurality of nodes and has a certain topology but is unknown, the sensing network is composed of sensing equipment and analyzing equipment, and the sensing equipment senses the target network and performs data preprocessing, and then transmits the target network to the analyzing equipment for topology inference.
Referring to fig. 2, the topology sensing method for a non-cooperative wireless network according to the present invention performs correlation detection on a node according to a signal sending time and a sending node serial number, and analyzes a communication relationship of a target network to obtain a topology structure, specifically including:
step 1: acquiring a node serial number and a sending moment of data sent by a target network within a period of time through a signal detection mechanism to form a data matrix P;
step 2: performing Glan's cause and effect hypothesis test on the data matrixes pairwise, solving the mean value of Glan's cause and effect zero distribution by using a time window method, and screening to obtain a potential neighbor set of the nodes as a threshold value;
and step 3: and carrying out conditional granger causal hypothesis test according to the data matrix and the potential neighbor set group, solving the mean value of conditional granger causal zero distribution by using a time window method, and screening as a threshold to obtain a final neighbor set of the node to realize topology perception.
Referring to fig. 3, the step 1 specifically includes:
the node sequence number and the transmission time are embodied in a data matrix P, which is a 0, 1 matrix. Such as if x in the data matrix P1,2If 1, then a packet is sent at time 2 on behalf of the node with node sequence number 1.
The sensor compares the received signal energy value with a signal detection threshold by sampling to determine if a signal is present:
if the signal exists, discretizing the signal according to the time slot by taking the falling edge of the signal as a standard, as shown in the following formula,
Figure BDA0003084445310000061
wherein xi,lData value, T, representing the ith time slot of the ith nodesIndicating the time length of the time slot, tiRepresenting the signal arrival time of the ith node;
thereby obtaining a preprocessed data matrix P
Figure BDA0003084445310000071
Step 2, performing the glange causal hypothesis test on the data in the data matrix in pairs according to rows, wherein the glange causal hypothesis test comprises setting a zero hypothesis and a candidate hypothesis and calculating a statistical test quantity, and the specific steps are as follows:
the data matrix P is the sum of all variables in steps 2.1.1-2.1.2, one variable for each row of P.
Step 2.1.1: data was modeled as the grand causal hypothesis:
for variables X and Y, zero hypothesis H0：
Assuming that without prior information for the variable Y, the information for X can be well predicted, i.e. Y cannot cause X by glange;
alternative hypothesis H1：
Assuming that the a priori information of Y helps the prediction of the information of X, i.e. Y granger causes X, it is expressed by the following equation:
Figure BDA0003084445310000072
Figure BDA0003084445310000073
wherein X [ n ]]Is the value of X in the nth slot, aiAnd biIs a parameter of linear regression, quantifies the degree of influence of past values on the current value, k is the order of the model, ε [ n ]]And eta n]Are each H0And H1An error of (2);
step 2.1.2: calculating a statistical test quantity G:
Figure BDA0003084445310000074
wherein
Figure BDA0003084445310000075
Denotes a null hypothesis H0The sum of the squares of the residuals of (c),
Figure BDA0003084445310000076
representing alternative hypothesis H1The sum of the squared residuals of (c). k denotes the order of the model, M denotes the number of observation total slots, F (d)1,d2) Denotes F distribution, d1＝k，d2＝M-2k-1。
Step 2, calculating the mean value of the granger causal zero distribution by using the time window method, and screening to obtain a potential neighbor set of the node as a threshold, wherein the method specifically comprises the following steps:
step 2.2.1: selecting a proper time window number nwE.g. nwWhen data is used 2The intermediate window is divided into nwThen calculating the granger causal test statistic G;
for example, for variables X and Y, they are classified as X1，X2，Y1，Y2Then separately calculate X2For Y1、Y2To X1The test statistic G of (a);
step 2.2.2: calculating the test statistic of all nodes, then calculating the mean value, screening the test statistic, considering that the two variables are related when the mean value is larger than the threshold value, and considering that the two variables are unrelated when the mean value is not larger than the threshold value; the relevant variables are then put together as a set of potential neighbors for the node.
And 3, performing conditional grand cause and effect hypothesis test according to the data matrix and the potential neighbor set group, wherein the specific steps are as follows:
each row of data of the data matrix corresponds to one variable, and each variable corresponds to one potential neighbor set. In the formula of 3.1.1, X is a variable taken from the data matrix, L, Y is a variable taken from the set of potential neighbors K, which are then examined by the conditional Grave hypothesis.
Step 3.1.1: in a neighbor set K of a node X, respectively checking whether each node has a special effect on the information prediction of the node X;
null hypothesis S0: all node information except Y in the known K set plays a role in predicting node X information, namely Y plays no special role in X;
alternative hypothesis S1: compared with the null hypothesis, the prior information of Y has a role in predicting X information, namely Y has a special role in X and is expressed by the following formula:
Figure BDA0003084445310000081
Figure BDA0003084445310000082
wherein X [ n ]]Is the value of X in the nth slot, ai,biAnd eiIs a parameter of linear regression, so as to quantify the degree of influence of the past value on the present value, k is the order of the model, ε [ n ]]And eta n]Are each H0And H1An error of (2);
step 3.1.2: calculating the statistical test quantity Gc：
Figure BDA0003084445310000091
Wherein
Figure BDA0003084445310000092
Denotes a null hypothesis H0The sum of the squares of the residuals of (c),
Figure BDA0003084445310000093
representing alternative hypothesis H1The sum of the squared residuals of (c). k denotes the order of the model, M denotes the number of observation total slots, F (d)1,d2) Denotes F distribution, d1＝k，d2＝M-2k-1。
And 3, solving the mean value of conditional grande causal zero distribution by using a time window method, and screening to obtain a final neighbor set of the node as a threshold, wherein the method specifically comprises the following steps:
step 3.2.1: selecting a suitable number of time windows, e.g. nwData were divided into two parts with a time window and the conditional granger causal test statistic G was calculatedc. For example, for variables X and Y, they are classified as X1，X2，Y1，Y2Then separately calculate X2For Y1、Y2To X1Test statistic Gc；
Step 3.2.2: calculating the test statistic of all nodes, then calculating the mean value, screening the test statistic, considering that the two variables are related when the mean value is larger than the threshold value, and considering that the two variables are unrelated when the mean value is not larger than the threshold value; the relevant variables are then put together as a set of potential neighbors for the node. Referring to fig. 4, the topology sensing apparatus facing a non-cooperative wireless network of the present invention includes:
a data preprocessing module: the data matrix P is used for preprocessing the sensed data so as to obtain a data matrix P as initial data of topology inference;
a potential neighbor identification module: performing the granger causal hypothesis test by using the data provided by the data preprocessing module, calculating test statistic, and performing potential neighbor screening by using the mean value of zero distribution as a threshold value;
and a final neighbor identification module: and carrying out conditional granger causal hypothesis test by using the processing result of the potential neighbor identification module, calculating statistic, and carrying out final neighbor screening by using the mean value of zero distribution as a threshold value.
The data preprocessing module comprises:
an initialization unit: initializing the device and emptying the data in the previous time period;
a sensing unit: the method comprises the steps of sensing signals, and obtaining signal sending time and node serial numbers;
a processing unit: the signal discretization method comprises the steps of discretizing a signal by taking a signal falling edge as a standard;
a storage unit: the data processing device is used for storing the processed data;
an output unit: for outputting the processed data to the potential neighbor identification module and the final neighbor identification module.
The potential neighbor identification module and the final neighbor identification module both comprise:
statistic calculation unit: test statistics for calculating a granger causal hypothesis test;
a zero distribution calculation unit: for calculating the mean of the granger causal zero distribution;
the potential neighbor identification module further comprises: potential neighbor screening unit: for screening a set of potential neighbors;
the final neighbor identification module further comprises: and a final neighbor screening unit: for screening the final set of neighbors.
Referring to fig. 5, the topology awareness system facing the non-cooperative wireless network of the present invention includes:
a memory for storing instructions and data;
a processor coupled to the memory, the processor configured to invoke and execute instructions and data stored in the memory, specifically:
the processor is used for preprocessing the sensed data to obtain a data matrix P which is used as initial data of topology inference; the method is used for carrying out the Glange causal hypothesis test, calculating test statistics, and carrying out potential neighbor screening by taking the mean value of zero distribution as a threshold value; the method is used for carrying out conditional grand causal hypothesis test, calculating statistic, carrying out final neighbor screening by taking the mean value of zero distribution as a threshold value, and storing the threshold value in a memory.
The topology awareness apparatus and system oriented to the non-cooperative network in this embodiment may execute the technical solution of the embodiment of the method shown in fig. 3, and the implementation principle thereof is similar, and will not be described herein again.
In the invention, the topology perception problem of the non-cooperative wireless network with the background and the characteristics of the embodiment is solved through the topology perception algorithm based on the conditional granger cause and effect, so that the communication relation of the target unknown network is obtained, the topology structure of the target unknown network is deduced, and a foundation is laid for the subsequent task.
The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above-mentioned embodiments, and all technical solutions belonging to the idea of the present invention belong to the protection scope of the present invention. It should be noted that modifications and embellishments within the scope of the invention may be made by those skilled in the art without departing from the principle of the invention.

Claims (10)

1. A topology sensing method facing a non-cooperative wireless network is characterized in that according to the sending time of a signal and the serial number of a sending node, correlation detection is carried out on the node, the communication relation of a target network is analyzed, and therefore a topology structure is obtained, and the method specifically comprises the following steps:
step 1: acquiring a node serial number and a sending moment of data sent by a target network within a period of time through a signal detection mechanism to form a data matrix;
step 2: performing Glan's cause and effect hypothesis test on the data matrixes pairwise, solving the mean value of Glan's cause and effect zero distribution by using a time window method, and screening to obtain a potential neighbor set of the nodes as a threshold value;
and step 3: and carrying out conditional granger causal hypothesis test according to the data matrix and the potential neighbor set group, solving the mean value of conditional granger causal zero distribution by using a time window method, and screening as a threshold to obtain a final neighbor set of the node to realize topology perception.
2. The topology awareness method for a non-cooperative wireless network according to claim 1, wherein the step 1 specifically comprises:
the sensor compares the received signal energy value with a signal detection threshold by sampling to determine if a signal is present:
if the signal exists, discretizing the signal according to the time slot by taking the falling edge of the signal as a standard, as shown in the following formula,
Figure FDA0003084445300000011
wherein xi,lData value, T, representing the ith time slot of the ith nodesIndicating the time length of the time slot, tiRepresenting the signal arrival time of the ith node;
thereby obtaining a preprocessed data matrix P
Figure FDA0003084445300000012
3. The topology sensing method oriented to the non-cooperative wireless network according to claim 1, wherein the step 2 of performing the granger causal hypothesis test on the data in the data matrix two by two rows includes setting a null hypothesis and a candidate hypothesis and calculating a statistical test quantity, and specifically includes the following steps:
step 2.1.1: data was modeled as the grand causal hypothesis:
for variables X and Y, zero hypothesis H0：
Assuming that without prior information for the variable Y, the information for X can be well predicted, i.e. Y cannot cause X by glange;
alternative hypothesis H1：
Assuming that the a priori information of Y helps the prediction of the information of X, i.e. Y granger causes X, it is expressed by the following equation:
Figure FDA0003084445300000021
Figure FDA0003084445300000022
wherein X [ n ]]Is the value of X in the nth slot, aiAnd biIs a parameter of linear regression, quantifies the degree of influence of past values on the current value, k is the order of the model, ε [ n ]]And eta n]Are each H0And H1An error of (2);
step 2.1.2: calculating a statistical test quantity G:
Figure FDA0003084445300000023
wherein
Figure FDA0003084445300000024
Denotes a null hypothesis H0The sum of the squares of the residuals of (c),
Figure FDA0003084445300000025
representing alternative hypothesis H1The sum of the squared residuals of (c). k denotes the order of the model, M denotes the number of observation total slots, F (d)1,d2) Denotes F distribution, d1＝k，d2＝M-2k-1。
4. The topology sensing method for the non-cooperative wireless network according to claim 3, wherein the method using the time window in step 2 finds a mean value of the grand causal zero distribution, and obtains a potential neighbor set of the node as a threshold screening, specifically as follows:
step 2.2.1: selecting a proper time window number nwDividing the data into n with a time windowwThen calculating the granger causal test statistic G;
step 2.2.2: calculating the test statistic of all nodes, then calculating the mean value, screening the test statistic, considering that the two variables are related when the mean value is larger than the threshold value, and considering that the two variables are unrelated when the mean value is not larger than the threshold value; the relevant variables are then put together as a set of potential neighbors for the node.
5. The topology sensing method oriented to the non-cooperative wireless network according to claim 1, wherein the step 3 of performing conditional grand cause and effect hypothesis test according to the data matrix and the potential neighbor set group includes the following specific steps:
step 3.1.1: in a neighbor set K of a node X, respectively checking whether each node has a special effect on the information prediction of the node X;
null hypothesis S0: all node information except Y in the known K set plays a role in predicting node X information, namely Y plays no special role in X;
alternative hypothesis S1: compared with the null hypothesis, the prior information of Y has a role in predicting X information, namely Y has a special role in X and is expressed by the following formula:
Figure FDA0003084445300000031
wherein X [ n ]]Is the value of X in the nth slot, ai,biAnd eiIs a parameter of linear regression, so as to quantify the degree of influence of the past value on the present value, k is the order of the model, ε [ n ]]And eta n]Are each H0And H1An error of (2);
step 3.1.2: calculating the statistical test quantity Gc：
Figure FDA0003084445300000032
Wherein
Figure FDA0003084445300000033
Denotes a null hypothesis H0The sum of the squares of the residuals of (c),
Figure FDA0003084445300000034
representing alternative hypothesis H1The sum of the squared residuals of (c). k denotes the order of the model, M denotes the number of observation total slots, F (d)1,d2) Denotes F distribution, d1＝k，d2＝M-2k-1。
6. The topology sensing method for the non-cooperative wireless network according to claim 5, wherein the method using the time window in step 3 finds the mean of conditional grand causal zero distributions, which is used as a threshold to obtain a final neighbor set of the node, and the specific steps are as follows:
step 3.2.1: selecting a suitable number of time windows, e.g. nwData were divided into two parts with a time window and the conditional granger causal test statistic G was calculatedc. For example, for variables X and Y, they are classified as X1，X2，Y1，Y2Then separately calculate X2For Y1、Y2To X1Test statistic Gc；
Step 3.2.2: calculating the test statistic of all nodes, then calculating the mean value, screening the test statistic, considering that the two variables are related when the mean value is larger than the threshold value, and considering that the two variables are unrelated when the mean value is not larger than the threshold value; the relevant variables are then put together as a set of potential neighbors for the node.
7. A topology aware apparatus for non-cooperative wireless networks, comprising:
a data preprocessing module: the data matrix P is used for preprocessing the sensed data so as to obtain a data matrix P as initial data of topology inference;
a potential neighbor identification module: performing the granger causal hypothesis test by using the data provided by the data preprocessing module, calculating test statistic, and performing potential neighbor screening by using the mean value of zero distribution as a threshold value;
and a final neighbor identification module: and carrying out conditional granger causal hypothesis test by using the processing result of the potential neighbor identification module, calculating statistic, and carrying out final neighbor screening by using the mean value of zero distribution as a threshold value.
8. The topology aware apparatus oriented to a non-cooperative wireless network according to claim 7, wherein the data preprocessing module comprises:
an initialization unit: initializing the device and emptying the data in the previous time period;
a sensing unit: the method comprises the steps of sensing signals, and obtaining signal sending time and node serial numbers;
a processing unit: the signal discretization method comprises the steps of discretizing a signal by taking a signal falling edge as a standard;
a storage unit: the data processing device is used for storing the processed data;
an output unit: for outputting the processed data to the potential neighbor identification module and the final neighbor identification module.
9. The topology awareness apparatus facing a non-cooperative wireless network according to claim 7, wherein the potential neighbor identification module and the final neighbor identification module each comprise:
statistic calculation unit: test statistics for calculating a granger causal hypothesis test;
a zero distribution calculation unit: for calculating the mean of the granger causal zero distribution;
the potential neighbor identification module further comprises: potential neighbor screening unit: for screening a set of potential neighbors;
the final neighbor identification module further comprises: and a final neighbor screening unit: for screening the final set of neighbors.
10. A topology aware system for non-cooperative wireless networks, comprising:
a memory for storing instructions and data;
a processor coupled to the memory, the processor configured to invoke and execute instructions and data stored in the memory, specifically:
the processor is used for preprocessing the sensed data to obtain a data matrix P which is used as initial data of topology inference; the method is used for carrying out the Glange causal hypothesis test, calculating test statistics, and carrying out potential neighbor screening by taking the mean value of zero distribution as a threshold value; the method is used for carrying out conditional grand causal hypothesis test, calculating statistic, carrying out final neighbor screening by taking the mean value of zero distribution as a threshold value, and storing the threshold value in a memory.
CN202110576245.9A 2021-05-26 2021-05-26 Topology sensing method, device and system for non-cooperative wireless network Active CN113328881B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110576245.9A CN113328881B (en) 2021-05-26 2021-05-26 Topology sensing method, device and system for non-cooperative wireless network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110576245.9A CN113328881B (en) 2021-05-26 2021-05-26 Topology sensing method, device and system for non-cooperative wireless network

Publications (2)

Publication Number Publication Date
CN113328881A true CN113328881A (en) 2021-08-31
CN113328881B CN113328881B (en) 2022-05-03

Family

ID=77416833

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110576245.9A Active CN113328881B (en) 2021-05-26 2021-05-26 Topology sensing method, device and system for non-cooperative wireless network

Country Status (1)

Country Link
CN (1) CN113328881B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114647198A (en) * 2022-03-09 2022-06-21 深圳市经纬纵横科技有限公司 Intelligent home control method and system based on Internet of things and electronic equipment

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106773693A (en) * 2016-12-21 2017-05-31 浙江大学 A kind of sparse causality analysis method of Industry Control multi-loop oscillation behavior
CN108684051A (en) * 2018-05-11 2018-10-19 广东南方通信建设有限公司 A kind of wireless network performance optimization method, electronic equipment and storage medium based on cause and effect diagnosis
CN109407654A (en) * 2018-12-20 2019-03-01 浙江大学 A kind of non-linear causality analysis method of industrial data based on sparse depth neural network
CN110855478A (en) * 2019-10-30 2020-02-28 南京航空航天大学 Single-sensor topology sensing method and device for unreliable information
WO2020046261A1 (en) * 2018-08-27 2020-03-05 Siemens Aktiengesellschaft Systematic prognostic analysis with dynamic causal model

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106773693A (en) * 2016-12-21 2017-05-31 浙江大学 A kind of sparse causality analysis method of Industry Control multi-loop oscillation behavior
CN108684051A (en) * 2018-05-11 2018-10-19 广东南方通信建设有限公司 A kind of wireless network performance optimization method, electronic equipment and storage medium based on cause and effect diagnosis
WO2020046261A1 (en) * 2018-08-27 2020-03-05 Siemens Aktiengesellschaft Systematic prognostic analysis with dynamic causal model
CN109407654A (en) * 2018-12-20 2019-03-01 浙江大学 A kind of non-linear causality analysis method of industrial data based on sparse depth neural network
CN110855478A (en) * 2019-10-30 2020-02-28 南京航空航天大学 Single-sensor topology sensing method and device for unreliable information

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
PAUL TILGHMAN; DAVID ROSENBLUTH: "Inferring Wireless Communications Links and Network Topology from Externals Using Granger Causality", << MILCOM 2013 - 2013 IEEE MILITARY COMMUNICATIONS CONFERENCE >> *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114647198A (en) * 2022-03-09 2022-06-21 深圳市经纬纵横科技有限公司 Intelligent home control method and system based on Internet of things and electronic equipment

Also Published As

Publication number Publication date
CN113328881B (en) 2022-05-03

Similar Documents

Publication Publication Date Title
US10929722B2 (en) Anomaly detection in streaming networks
CN108601026B (en) Perception data error attack detection method based on random sampling consistency
CN112165485A (en) Intelligent prediction method for large-scale network security situation
Yan et al. A hybrid intrusion detection system of cluster-based wireless sensor networks
CN112966714B (en) Edge time sequence data anomaly detection and network programmable control method
CN110855478A (en) Single-sensor topology sensing method and device for unreliable information
CN114143037A (en) Malicious encrypted channel detection method based on process behavior analysis
CN116684878B (en) 5G information transmission data safety monitoring system
CN117041019B (en) Log analysis method, device and storage medium of content delivery network CDN
Martins et al. A support vector machine based technique for online detection of outliers in transient time series
CN113328881B (en) Topology sensing method, device and system for non-cooperative wireless network
Sun et al. A wireless sensor network node fault diagnosis model based on belief rule base with power set
CN114048546A (en) Graph convolution network and unsupervised domain self-adaptive prediction method for residual service life of aircraft engine
CN115473748B (en) DDoS attack classification detection method, device and equipment based on BiLSTM-ELM
CN115987643A (en) Industrial control network intrusion detection method based on LSTM and SDN
YR et al. Iot streaming data outlier detection and sensor data aggregation
Tu et al. Intrusion detection based on generative adversarial network of reinforcement learning strategy for wireless sensor networks
Balakin et al. Detection of computer attacks using outliner method
Chow et al. FLARE: Detection and Mitigation of Concept Drift for Federated Learning based IoT Deployments
Qiu et al. Abnormal Traffic Detection Method of Internet of Things Based on Deep Learning in Edge Computing Environment
Jose et al. Prediction of Network Attacks Using Supervised Machine Learning Algorithm
CN113627621B (en) Active learning method for optical network transmission quality regression estimation
CN112565200B (en) Industrial control network misuse intrusion detection early warning system based on edge intelligence
CN117421684B (en) Abnormal data monitoring and analyzing method based on data mining and neural network
CN116541794B (en) Sensor data anomaly detection method based on self-adaptive graph annotation network

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant